Table 2 Specification of regression models.

From: Federal criminal sentencing: race-based disparate impact and differential treatment in judicial districts

	Regression model
	1	2	3	4	5	6	7	8	9	10	11	12
NEWRACE	⬤	⬤	⬤	⬤	⬤	⬤	⬤	⬤	⬤	⬤	⬤	⬤
AGE		⬤	⬤	⬤	⬤	⬤	⬤	⬤	⬤	⬤	⬤	⬤
MONSEX		⬤	⬤	⬤	⬤	⬤	⬤	⬤	⬤	⬤	⬤	⬤
NEWEDUC		⬤	⬤	⬤	⬤	⬤	⬤	⬤	⬤	⬤	⬤	⬤
SENTYR			⬤	⬤	⬤	⬤	⬤	⬤	⬤	⬤	⬤	⬤
NEWCNVTN				⬤	⬤	⬤	⬤	⬤	⬤	⬤	⬤	⬤
XCRHISSR					⬤	⬤	⬤	⬤
CHAP2						⬤
COADJLEV							⬤
XFOLSOR								⬤
GRID (XCRHISSR:XFOLSOR)									⬤	⬤	⬤	⬤
MANDMIN										⬤	⬤	⬤
GOVDEPART											⬤	⬤
CIRCDIST												⬤

The results summarized in Fig. 1 and Table 1 come from these 12 linear regressions. For each model, the response variable is the length of the sentence, SENTTOT. The explanatory variables included in each model are those marked with a bullet in the table below. See “Methodology” for a discussion of variable meanings.

Back to article page

Search

Advanced search

Quick links